Land Reform: Summaries of Pertinent Articles On South African Land Reform Simon Halliday April 2007

1

Fenwick and Lyne (1999)

1.1

Areas

The researchers used two areas in KwaZulu-Natal for a total of 150 households. 1. Mpembeni Ward: Broken topography, unreliable rainfall, household consumption crops, with local sales of maize and vegetables. 2. Mkhwanazi Ward: flat topography, high rainfall, poor soil, sugar cane as cash crop.

Mean Annual Cash Income Non-farm Cash Received Commercial Bank Savings Cattle Small Livestock HH Head Eduation Distance from Formal market Migrant Workers

Mpembeni R323

Mkhwanazi R971

R16 013

R15 825

1658

1285

6 15 4

3 8 5

>83

<25

1,2

1,5

Table 1: Mean Values of Pertinent Stats As a consequence of the above, transaction costs are expected to be higher in Mpembeni. 1

1.2

Empirical Analysis

Small scale farming is constrained by factors other than liquidity such as: 1. Imperfect land markets 2. Inadequate family labour (quality and quantity) 3. Poor information 4. High transaction costs Dependent variable developed using: 1. expenditure on seasonal farm inputs ≥ R800 2. ownership of farm assets 3. adoption of sugar cane/timber enterprises 4. crop sales >R300. Thus the sample was broken into farmers (1) and non-farmers (0) contingent on the conditions above. They used Logit MLE. Farm income excluded as explanatory because it was used to define dependent var. Remittances, pensions and microenterprise income were used.Family labour was used as a variable (ratio of family size, and squared), as was an index of transaction cost values, transaction costs were higher in Mpembeni than in Mkhwanazi. Considering pertinent descriptive statistics: non-farmers much less liquid than farmers (11 947 against 36 678); both equally held informal savings, famers held slightly more formal savings, farmers better endowed with family labour (although both groups’ dependency ratios were similar), The model that was estimated predicted 85% of cases correctly. The model had reduced classification error. The signs on the variables considered (Farm Size, Rent, Liquidity, Family Savings, Family Labour, Visits) all held to a priori expectations of positivity. Shows that poor farmers are constrained by low incomes and savings, inadeqaute information, small farm sized, all of which is compunded by insecure land tenure and risks associated with rental transactions and finally high transaction costs.

1.3

Conclusions

Small farmers constrained by low levels of liquidity. Farm size was also a constraint - farmers with larger farms more likley to farm intensively. Hence if there was a rental market for land-poor farmers they could increase the scales of their production. Poor information access (i.e. fewer visits 2

from extension officers) and high transaction costs are important. Thus improving access to information and decreasing transaction costs could lead to the alleviation of these constraints.

2

Graham and Lyne (1999)

General: Authors take the deeds registry of 1997 and assess these with respect to quantity of lanf, quality of land, characteristics of owner and several other salient factors.

2.1

Modes of Land Redistibution

Three principal modes of land redistribution 1. Government assisted land transfers: main govt. grant was a R15 000 cash grant to purchase and develop farmland. Problem: means test suposedly means that many credit-worthy applicants who would use the grant to leverage a loan would not be able to obtain the grant, and those who could obtain the grant would not be credit-worthy. 2. Private land purchases: Land transactions facilitated by the KwaZulu Finance and Investment Committee (KFC). Illovo sugar offered a system by which it would invest 18% of the purchase price with Ithala bank, thus basically discounting the price of its land by 18%. KFC then used this to reduce mortgage rates. The scheme financed some 90 disadvantaged sugar cane farmers with total land sales being roughly R80 million. 3. Equity share schemes: Share of equity between a commercial farmer and their disadvantaged employees. Most of these are in the WC involving capital intensive work. Not prevalent in KZN.

2.2

Data

Used deeds register for all KZN-based transactions. Total of 4737 transactions, of which 2 111 urban and 2626 rural. Took rural transactions for land acquired by one owner (plus spouse). Dropped transactions for residential development or those that transferred ownership from one disdvantaged individual to another, unless gender of owner changed in which case it was kept. Yielded a total of 1142 farmland transfers. This separated into two groups ’disdvantaged’ (183) and ’white’ (959). Gender not established on register - hence imputed via names or through source of mortgage loans. Magisterial district was identified.

3

2.3

Quality of Land

Used the Lyne and Darroch (1997) estimate of 5,3 million ha of area available for redistribution. From data, total area registered in 1997 was 372 995 ha, or 7% of the 5.3 mil. ha total. Transfers to disadvantged individuals totalled 22 934 ha. representing 6.2% of total land transferred during 1997, or 0.43% of total estimated redistributable land. This is much larger than the 0.09% estimate from Lyne and Darroch (1997) for 1995. Hence massive growth approximating 117% from 1995 till year end 1997.

2.4

Quality of Land

Price of land (per ha.) used as a proxy for quality of that land. Total farm size and thus total price were lower for disadvantaged farmers. Quality of land bought by disadvantaged entrants (R2302) slightly higher than non(R2103), but this may just a negative relationship between total farm size purchased and ha. price. Distribution of disadvantaged farmers bimodal more belwo R2500 than in the R10000 - R13000 interval. This indicates a quality differential.

2.5

Modes of Redistribution

Government assisted transfers (21 of total) had an average price per ha. of R1119. This indicates poor quality. Transactions financed with mortgage loans had an average price per ha. ≥R4900 indicating superior land quality. Cash transactions without govt. grants or mortgage loans were predominantly for relatively small farms, with a price per ha. of R1504 on average. Inheritance transactions did not have a price per hectare, transferred to approximately 1210 ha to disadvantaged individuals.

2.6

Spatial Distribution

Most mortgage-financed transactions occurred in high quality coastal belt, most cash purchases and all govt.-assisted transactions occurred in regions of lower quality farmland (lowveld and midlands).

2.7

Conclusions

On average quality of land the same for whites and disadvantaged, but with different modes of redistribution the quality of land was marked, similarly to when the spatial distribution was assessed. Women are under-represented except in inheritance transactions.

4

3 3.1

Lyne et al. (1996) Renting as a measure of efficient land use

Customary land tenure provides social security. In the absence of high transaction costs and uncertainty about guture income streams, incentives exist for the less productive farmer to sell to a more productive farmer. Where farms are small and land is valued for social security, there is a problem in price signaling - the reservation price of the owner may exceed the offer price of the potential productive farmer - hence creating problems for land sale measures of productivity. Problems with moral hazard and transaction costs can decrease incentives to invest in rented land.

3.2

Constrained Rental Markets

A highly differentiated wage labour market indicating that HHs face very different economic opportunities (Thomson and Lyne, 1991). Risk and high transaction costs raise the reservation price of potential lessors and tent to confine the populatio ntof tenants to farmers who can cover the risk premiums charged by lessors. (13) In KZ renting is almost synonymous with intensive cropping and surplus production, even if lessors rent to family/friends to reduce risk. In a Transkei survey, where some perceived renting as ’Not Allowed’ those that rent-in land were more productive.

3.3

Rental Market Activity

Understand that group of observed tenants is likely to be characterised by famers confident of ability to cover premiums charged by lessors who risk losing land rights from leasing out land. Hence an inverse relationship between rental market activity and the productivity gap could exist. Even in KZ where market is particularly weak, renters have higher productivity when deducting additional costs. Renting also seems to transfer land to households that have more capital and labour per unit area. Area of land rented-in was found to be a very important determinant for farmers wishing to be emerging farmers (i.e. sell cash products). Supposedly rental would increase equity in KZ. But both lessors and lessees (potential and existent) are constrained by riskiness of market. Renting also helps by maintaining human and financial capital because it [rental] permits adjustment in the scale of farming operations in an environment where farms are diminishing in size.

3.4

Causes of inefficient land rental markets

For an efficient land market to exist there should be secure tenure and low transaction costs. In SA we should add that customary tenure is only

5

secure when it refers to the ability to use land for a certain period and for a define purpose without disturbance, this may change when the holder pursues transactions for their own interests. Security of tenure therefore has breadth, duration and assurance components - if assurance is lacking use rights become uncertain hence jeopardising possible transfers. Insecure tenure thus jeopardises the rental market. With respect to transaction costs, we observe that TCs vary inversely with security of tenure. TCs increase with poorly defined and allocated property rights. Risks that reduce tenure security can increase TCs (undermine assurance). Uncertainty about institutions and laws also increase TCs (because of increased risk). In KZ many disputes settled by a Tribal Authority. Hence, land rights that are locally administered and it appears that the procedures for establishing rental contracts are not clear - hence further decreasing likelihood of rental. A cross-section of countries showed that perceived riskiness of transactions or the subjection of transactions to customary restrictions showed an inverse relationship with incidence of renting. Rental markets are more active where: a. procedures for establishing contracts are transparent b. local precedents set in land disputes confirm security of tenure c. national law sanctions local precedents But problems: TC approach ignores issues of collective action. Similarly, a Marxist approach ignores Collective Action.

3.5

Policy Implications

Customary rights are seldom documented - hence ’replacement’ policies could create conflicting rights to land. Thus need adaptive policy. For an adaptive approach, need to identify ‘losers’ and compensate them using social security (pensions, unemployment benefits, etc), or options to exchange use rights for serviced residential sites. Also offering incentives to tribal authorities may be necessary - empirically those authorities that have been remunerated are more likely to endorse written contracts. Moreover, recording of transactions, disputes and precedents is a requirement for the adaptive model to work.

3.6

Conclusions

Renting closes productivity gap by transferring land to more efficient users. Has equity advantages as it avoids distress sales. But this is constrained by customary tenure being insecure and high transaction costs. There is a suggestion for support programs that encourage farmers to pressurise for 6

exclusive land rights, but these ignore collective action problems and reactionary resistance from other groups. Adaptive programs are proposed as the best path towards reform.

4

Williams (1996)

General Comment: this article is a critique of the World Bank report, Options for Land Reform and Rural Restructuring in South Africa (1993) (hereafter Options, in addition to offering criticisms of two papers by ‘South African Land Policy’ and, with Feder, Power, Distortions, Revolt and Reform in Agricultural Land Relations (Binswanger et al., 1993). Options came out of several phases on interaction between the World Bank and SA policymakers, including the Land and Agricultural Policy Centre (LAPC), towards a contribution to the Rural Restructuring Programme (RRP). Williams’ aim is to critique the underlying ideological and institutional presupposition of the report.

4.1

Social Engineering in Kenya and South Africa

Includes a comparison and contrast of situations in Kenya, SA and Zimbabwe. Kenya as positive, Zim as negative. Similarities pre-1960 include: segregation of land ownership, suppor for large-scale farmers through subsidy (direct and non-), marketing boards, exclusion of smallholders, measures for labour extraction from tenants, policy encouragement for farmers to be established on native lands involving shift from communal tenure to individual title. What follows in the text was the reinforcement of the factual history of the above similarities. Importantly, Williams notes the Tomlinson Commission(SA) and the Swynnertwon Report in Kenya (1954), both of which sought to promote commercail production by a class of ’progressive’ farmers. Williams also discusses the similarity between the conditions that facilitate livingin Kenya that are similar to SA: families depend on remittance income, male migrants prefer not to provide reousrce to wives to invest in farming when they cannot control the investment process, survival necessitates multi-pronged strategies combining farm and non-farm income. Centrally about ’juggling survival’ rather than the accumulation of wealth. In attempts to develop rural SA, the government attempted to use a centrally managed, capital intensive farming system, but it was too expensive. Teh Development Bank of South Africa (DBSA) shifted its activities to the Farmer Support Programmes (FSP) modelled on WB integrated agricultural development projects. These had three categories1 that were remarkably similar to those from Swynnerton. The plan was developed to help the 1

commercial farmers, emergent farmers and subsistence farmers

7

FSP through supply of inputs, machinery, extension, credit and marketing. ‘Hardly any commercial farmers ‘emerged’ from the process’ (145). Which led to the opinion that credit, input and extension services don’t solve the land problem. In KwaZulu (KZ), although men and women operating on small parcels of land through customary tenure could augment family income, tensions between millers and growers resulted from credit and extension services. However there were additional positive results in KwaNgwane (KN), where there had been historical relocation into ’betterment villages’, with wage-labour-dependent farming. This shows how ‘a single processor can yield a higher return to producers than local alternatives’ (146).

4.2

The uses of History

What follows is Williams’ commentary on Binswanger et al. (1993). They argue that titling of land leads to transfers to more efficient farmers and the availability of credit markets, however there is a concomitant and consequent use of power relations to extract economic rents from land, labour and consumers. Williams argues that they fail to identify how the changes may not be an evolution towards a pinnacle of farming, but rather that different systems may be complementary to one another because of different forms of labour relations, deployment of skills, different categories of workers and so on. He claims that Binswanger et al. (1993) fail to see how many South Africans strove agains their subjugation as wage labour. Williams says that the Binswanger et al. (1993) argument for the existence and evolution of Junker Estates is produced in order to prescribe a policy that, in ’the absence of market imperfections and state support for mechanized farms’ (148), supports the ideal of small family farms. But with previous intense mechanization, there cannot be enough agrarian-skilled families to establish these small farms. Moreover, the establishment of such small-scale farms will necessitate new state interventions in the body of grant systems and additional support. Moreover, small farmers could sell back to big farmers once they have obtained land, hence there needs to be ‘a removal for all implicit and explicity suppor distortions favouring large farms’ (Binswanger et al., 1993, 62) - thus Williams argues that there is an implicit connection between Binswanger et al. (1993)’s arguments and the ideology (rhetoric) of liberalizing markets and eliminating subsidies.

4.3

Removing Distortions

The main argument in Binswanger et al. (1993, 2) is that ’small-scale farm models would tend to be more efficient than large-scale models when operating in an undistorted policy environment’. From which we can conclude that a more equitable distribution of land will result in greater efficiency of production. Williams counters saying that their argument rests on two

8

bases: one, that small-scale farming is necesssary more efficient than large wage-labour-dependent farming and there are few if any economies of scale in agriculture. Williams says that they argue, based on Bundy (1979), that the size of commercial farms was as much a measure to exclude black people from owning land, as it was to subsidise white acquisition of machinery and inputs as well as securing power over labour, rather than necessarily being as a distinct measure to increase efficient production. Williams refers to Moll (1988) and others who have evidenced the existence of economies of scale in SA farming, moreover scale efficiency seems to ‘scale efficient, depending on how farmers organize their specific variable and fized input mix, as well as the combination of outputs they produce.’ Moreover, smallholders who try to emulate larger farming practices ’likely to run into problems of high unit costs’. Williams further criticizes Binswanger et al because of their use of TFP as an approximation of farming efficiency from ?, the derivation of which Williams claims is questionable. Later world bank reports qualified earlier statements, saying that, operating in a distortion-free economy, small-scale farming would be at least as efficient as large-scale farming. The World Bank sponsored further work on liberalization, focusing on subsidies and tariffs. They conceded that there needs to be a commensurate response from the first world to remove its market distortions and claims that, if that were to occur, a more efficient market in South African agriculture would exist. The structure of the marketing board system also needed to be addressed, however only in a way that equally ensured the maintenance of levels of employment and output capacity. Williams argues that the removal of subsidies will not ’level the playing field’, he asserts that most African producers will be relatively disadvantaged by transport constraints: distance from railway, trains and major markets. With respect to imports, the World Bank’s argument is that ’South Africa could use impots to manage consumer prices for basic foods within acceptable limits’. However, the US and EU prices will always be below domestic prices on the coast because of their subsidy programs hence possibly resulting production problems domestically. Moreover, world prices are only really applicable to the coastal/high trade regions of South Africa. For more distant regions the benefits of lower prices would not be realised because of prohibitively high transport costs. Lastly, white maize, the main staple of most of South Africa’s population, is not produced internationally at volumes to satisfy the South African market. Hence, international trade liberalization may not provide the source of efficiency that the World Bank claims it should.

4.4

The Case for Land Reform

The main arguments for Land Reform rest on 1) the claims of people to return to land from which they were forcibly dispossessed, 2) the insecu9

rity and poverty of people living and working in, or evicted from, white farming areas, 3) continued majority poverty in former homelands. The World Bank argument was that there should be incentive compatible contributions by prospective tenants, but that this would therefore differentiate between the poor and non-poor because of this requirement. However, they believed that the new (black) land owners would produce benefits to the poor through tenancy or wage-work. Williams argues this is a return to the colonial argument progressed by Swynnerton. He highlights another argument, which discusses the problems of inequality, entry control mechanisms, and non-reliance on market-mechanisms for resource allocation. Williams contends that the claim that declines in profitability in certain sectors ‘seems to stymie any prospects of expanding agricultural production’ (157).

4.5

Redistributing Land

Problems of individual rights vs. communal rights. Allocations need to be made both according to poverty (need) and to managerial ability (productive capability). Because of these, and other, claims the World Bank’s assertions become ambiguous argues Williams. There are several areas that need to be considered: • Property Rights: Land claims for restitution against dispossession that took place after the Native Lands Act of 1913. Expropriation could be done ’for public purpose’ (roads) and ’in the public interest’ (redistribution) Sums paid in compensation could take into account history, market value, investment in land, interests of those affected, degree of previous state investment. The govt. limited expropriation to ’public purposes’, which could jeopardise the redistributive process. • Acquiring Land: Could use state land. Alternatively, eliminate policies favouring large-scale farming would bring prices down. Offer social security to owners, or through compellence by using banks to foreclose on unpaid debt. Subsequent to all of this could convene a committee to review ’expropriation in the public interest’. • Who would benefit?: Differentiation between those who would explicitly use the land productively and those for whom the focus would be on poverty/welfare alleviation. However, the implementation for such identification and subsequent policy facilitation (costs & testing) was not considered (or left out) by the World Bank Williams argues. Moreover, he claims this is a return to the colonial ideologies of Tomlinson w.r.t. the new ’emergent farmer’ nomenclature. Moreover, conflicting claims on land could create conflict (tenancy vs. restitution claims for example).

10

• What would it cost?: Williams: “We don’t know.” Estimates by WB say R3.5 billion, but based on models and assumptions that are problematic. One main issue: administration costs over five years estimated at R944 million, which Williams believes is totally insufficient and therefore misleading. Moreover, without public subsidies or guarantees many of the projects won’t work. • Who pays?: The taxpayer. WB report didn’t provide clarity on which donors, if any, would pay the R550 million per year to but out white farmers. Foreign borrowing could postpone payment of money, but would require interest and repayment in dollars. Even accepting the WB’s expected costs, R10 billion would have to be found in public funds over the first five years. Hence, the project would need to be scaled down because of indebtedness, which could jeopardise the project. WB insists it’s models are tentative and as drivers, Williams responds: ”If the models are not to be used in planning, and are not financially realistic, what are they for?’

4.6

Planning for the Future

Williams argues that the WB idea of ‘decentralization’ is actually just ‘development speak’ for ‘extended central control over rural people’ (163). He also shows frustration at the seeming rhetoric on community involvement in the policies which vacillates from a discussion of top-down management to ’participatory rural appraising’. However, when informed by Lund and Wakelin who argue that there should be elected district councils, the WB shifts terminology to a ’local development council’ (not elected, but under the control of the national body). However, Williams does not clarify whichis superior, he merely offers criticisms of what is said. A further complication that the World Bank avoids is that of the problem of the rhetoric associated with rural resettlement schemes. Williams argues that almost any characterization of land reform will involve rural resettlement. He claims that the WB avoids this because of the stigma attached to the concept and because of the often prohibitive costs associated with resettlement relative simply to reform that would target groups or individuals already living on land. Williams draws attention to the fact that the WB tries to solve the developmental dilemmas inherent in land reform in South Africa, buthe says that they are not resolved. Moreover, he claims that they do not establish rigorously whether their grand plan can be afforded. He asserts that they do not adequately consider the separation between ‘land reform’ as an ideal and the broader implementation issues that are confronted by rural people generally. His conclusion is that the World Bank plan to transfer 30% of land in 5 years is unrealistic and that its scaling down to the R15 000 grant

11

is unlikely to realise the World Bank’s vision. Williams concludes that land reform is always going to be a particular consideration and not at all general - different circumstances obtain differently. The World Bank does not take sufficient cognizance of this when attempting to create a uniform approach to South African land reform.

5

Zimmerman (2000)

Zimmerman elucidates on how the demand-led ‘willing-buyer’ system must inevitably exclude (large portions of) the most vulnerable (women, the poor and the landless) in the population, i.e. those who have the greatest need for programs geared toward their assistance. He argues that unless there are ancillary programs for credit, insurance, extension and basic services these vulnerable groups will not participate in the land reform program.

5.1

HH Heterogeneity

Zimmerman argues that there is a problem in how the policy is framed, it intends to target ’the landless poor and women’ but also those ’who already have a foothold in agriculture’. Moroever, there is a political demand for racial restructuring which can, and will, occur possibly without a concomitant alteration in the make-up of poor versus non-poor participation in the program. With a demand-led program, participants self-select into the program. In two ways this alleviates stress on government: it rations benefits to those who are able [to farm], as well as reducing costs by providing the grant to those that can pay bear some costs themselves. Assuming that the black rural population is homogenous and poor, govt does not need to articulate how it will assess the poverty-alleviation impacts - reason: if all rural black people are poor then by definition land reform is poverty-alleviating. If not, then it may not be. Meaningful poverty alleviation is not just about the return of capital assets, but also of other assets: human capital (health and education), water rights, cattle, and so on. For land redistribution to result in poverty alleviation it requires: 1. credit or grants for up-front costs 2. risk mediation 3. human capital formation 4. reticulation of electricity and water 5. information networks

12

5.2

Up-front costs

Costs include: 1. direct program participation costs 2. moving costs 3. costs of new equipment 4. application and search costs 5. necessary land improvements 6. opportunity costs of foregone activities Zimmerman provides several estimates of the costs above showing how it would be prohibitively expensive for poor households to obtain the grant and to start farming. He also indicates that those households that don’t have savings are less likely to be able to pay the up-front costs (estimated at R5400 in 1993 prices). He shows that approximately 40% of poor households are rationed out simply by the costs of land, even with the price of land reduced by 80% through grants. This means that even more poor households should be rationed out when including the additional up-front costs. Zimmerman understands that the offering of swear-equity could aid this process, but he asserts later that the landless poor and female-headed households are labour constrained and thus less able to supply this sweat equity.

5.3

Risk

Risk-aversion, and down-side risk-aversion in particular, seems to be more prevalent among the poor. There are several potential sources of risk: 1. uncertainty about profits 2. yield risk and input-output price risk 3. uncertainty on the availability and wages of non-agricultural employment 4. by accepting grant income may jeopardise remittance income Generally, there is low profitability in the agricultural sector and farmers are propped up by inefficient subsidies. Moreover, though it is argued that the highly mechanized white farms aren’t themselves efficient, this does not imply the converse that low-mechanized small farms will be efficient. Geography and climate furthermore influence riskiness. Many households also do not believe that agriculture will be the main source of income, hence the prospects for non-agricultural income are important. We do not know 13

how many, or even if, rural jobs will be created through govt rejuvenation policy, nor the rate at which they will be created if they are. In terms of remittances, there is a problem of moral hazard that can be perceived by a remitter: if things do not go well, the grant receiver still gets the grant in addition to the remittance that they obtained previously. Hence, they may choose to stop remitting.

5.4

Human Capital

The assumption that has been held is that the extension services provided to white farmers could be reallocated to new black farmers. However, there is a problem with this assumption: 600 000 smallholders are likely to require far more assistance than 20 000 established farmers. The same number of trainers cannot service this new total of farmers. Morevoer, because extension is cumulative it is likely that more visits per trainer will be required to assist these numbers. Furthermore, the type and quality of service that these smallholders will require will be different to that which serviced the white farmers previously because of size of land and relative experience levels. As an additional constraint, education and training has not been explicitly costed for by the DLA - which implies an underlying assumption of prior knowledge or experience of the behalf of the applicants which could act as a further segregator from poor applicants. As an adjunct to this, the evidence shows that human capital/training is as important for the success of an agricultural project as access to land is. Zimmerman establishes from SALDRU’s 1993 PSLSD that levels of agricultural experience from the former homelands is substantially low. Zimmerman also claims that there is a lack of farm managerial capability as a result of poor education that is endemic to much of the black population. There is also further evidence to suggest that most ’successful’ farmers: are male, had parents that were farmers, lived in white areas, to have been an entrepreneur before commencing farming and to have generally higher managerial capabilities (1448). All of this therefore indicates that human capital is neither evenly distributed across the population, and that they are definitely not likely to be poor, landless women.

5.5

Time poverty

That land will be beneficial and productive rurally as a consequence of rural land reform is based, to a large extent, on the assumption of surplus rural labour Zimmerman argues. Restoring assets to the poor should, with their surplus labour, allow more efficient outcomes. However, Zimmerman progresses by saying that this assumption of surplus labour is problematic. In fat many rural households are labour constrained he says moreover he claims that it is as a result of previous (and possibly current) government

14

neglect for rural institutions and services that icreases the labour component necessary for daily living. Land overutilisation and underutilisation occur side by side. He has several arguments as to why. Firstly, there are market imperfections that cause underutilisation. There is insufficient labour and land rental markets - hence households with insufficient (or inefficient) labour cannot hire more labour and those households that are more efficient cannot hire more (contiguous) land because of there being no land rental market. A second explanation, progressed by Wilson and Ramphele (1989), is that historic patterns of migration have left the homelands denuded of productive labour. There is a differential between rural and urban household consumer:worker ratios, but this does not explain away the labour shortage. Thirdly, lack of services and institutions in rural areas requires a cumbersome, labour-intensive form of everyday household maintenance - there is a substantial difference in wood/water procural time in rural (54.35 hrs/week) vs. urban (3.28 hrs/week) areas. This third explanation of ‘time-poverty’ will require significant government intervention for its resolution. Zimmerman further argues that, when receiving the land grant, households may have to choose to have a smaller plot with services (the installation for which they pay) or a larger plot without services. This is problematic. Moreover, many households do not have the predicted required household labour, in fact Zimmerman provides evidence to suggest that only 55% of former homeland households have sufficient household labour to meet required labour inputs. Combining this with the service constraint and it seems as though only one third of possible targeted recipients will be able to participate in the program. Zimmerman consider an argument some have put forward that land acquisition will result in return migration by working males. But if it may, there is no measure of how many men will return, nor whether they will have the requisite skills. Another factors is simply taste for work, it seems as though the ’taste’ for agricultural work is skewly distributed to the older populace, young surveyed individuals do not voice a preference for working in agriculture. Moreover there seems to exist an overall negative attitude towards land.

5.6

Geographical considerations

Unlike other social upliftment programs where ‘benefits move to beneficiaries’ land redistribution is unique in that ’beneficiaries must move to the benefit’. Zimmerman models how large the geographic distances of moves will be and shows that, given specific assumptions, there will be large movements from the ex-bantustans to the ‘summer grain areas’. Moreover many moves will constitute ‘socially significant’ moves away from social support networks, burial sites, known markets and definitely not within commuting distance from current homes. Thus the redistribution program will consti15

tute migration as much as it will redistribution. It is important to note that this does not bode well for the program itself - evidence suggests that successful programs involve individuals working on land that they know (Lipton and Lipton, 1993). Furthermore, moves would be longer for women, and the poor than they would be for men. Lastly, because successful black men serve as role models, their departure may serve to worsen the agricultural situation in the areas from which they depart, possibly with additional negative externalities.

5.7

Implications for Policy

Generally, the poverty alleviation implications of land reform has been oversold. The policy excludes many of the poor, it involves high risk to those who could adopt it and it will involve travel over long distances. Demand-led targeting will not alleviate policy Zimmerman argues. The policy requirements will be for a rural development program that occurs in conjunction with land reform to deal with problems of risk, human capital, services, time constraints, etc. Insurance markets and social security need to be addressed. The costs of the program need to be considered, possibly through the use of compulsory sales (in the public interest?). Lastly the provision of basic services: irrigation, potable water, electricity, schools, health care must be provided in the redistributed areas in order to incentivise movement and alleviate time constraints.

5.8

Conclusions

There is a tension between equity/effiiciency considerations and poverty alleviation considerations. They tend to be conflated in land reform but this is not an accurate representation of what land reform does or is. This conflation is problematic because many of the benefits could go to (already wealthy) black farmers for whom (through taxes) the land may be paid for by those poor we are told the policy is meant to benefit. Moreover, with the ’good’ farmers moving away from the homelands the benefits that they provided (as employers, or socially) would move with them and this could leave a gap in that market that would further damage the poor. In order for it to be a poverty-alleviation program the system of land reform needs to change - it does not adopt typical poor-targeting mechanims (queues, low wages, specific foods), but uses demand-led selection which rations the poor out of the program.

References Binswanger, Deininger, K., and Feder (1993). Power, distortions, revolt and reform in agricultural land relations. World Bank Report. 16

Bundy, C. (1979). The Rise and Fall of the South African Peasantry. London. Fenwick, L. J. and Lyne, M. C. (1999). The relative importance of liquidity and other constraints inhibiting the growth of small-scale farming in KwaZulu-Natal. Development Southern Africa, 16(1):141–156. Graham, A. W. and Lyne, M. C. (1999). Land redistribution in KwaZuluNatal: an analysis of farmland transactions in 1997. Development Southern Africa, 16(3):435–445. Lipton, M. and Lipton, M. (1993). Creating rural livelihoods: some lessons for South Africa from experiences elsewhere. World Development, 21(9):1515–1548. Lyne, M. and Darroch, M. (1997). Broadening access to land markets: financing emerging farmers in South Africa. Development Southern Africa, 14(4):561–8. Lyne, M. C., Thomson, D., and Ortmann, G. (1996). Institutional change to promote land rental markets in the developing regions of Southern Africa. Agrekon, 35(1):12–19. Moll, D. (1988). The Structure of Commercial Agriculture in South Africa: a quantitative approach to economies of scale, farm size change and technical change. PhD thesis, University of Oxford. Thomson, D. and Lyne, M. C. (1991). A land rental market for KwaZulu: Implications for farming efficiency. Agrekon, 30(2):287–290. Williams, G. (1996). Setting the agenda: A critique of the World Bank’s rural restructuring programme for South Africa. Journal of Southern African Studies, 22(1):139–166. Wilson, F. and Ramphele, M. (1989). Uprooting Poverty: the South African Challenge. Cape Town: David Philip. Zimmerman, F. (2000). Barriers to participation of the poor in South Africa’s land redistribution. World Development, 28(8):1439–1460.

17